Shuangqing Wei
Louisiana State University, LA·Engineering
Networks of devices, from sensors to cameras, are increasingly asked to learn and decide together, often under tight limits on data, bandwidth and privacy. These two projects tackle that challenge from two sides.
- Fields
Engineering
Computer Science
Mathematics & Statistics
- Topics
privacy-preserving machine learning
differential privacy
video privacy
- Grants
- 2 known grants, through Sep 2028
What they work on
- Generate a synthetic video dataset of tracheostomy-dependent infants to train privacy mechanisms, identity separation and recombination systems, and classifiers when real patient video is scarce.
- Compare multiple privacy designs for medical video, including face pixelation and adversarial generative mechanisms, on how well they preserve detection accuracy versus how strongly they hide identity.
- Extend differential privacy theory to moving images of people, validating the guarantees with specialized classifiers and feedback from medical personnel and caretakers.
- Develop a hierarchical distributed optimization framework that learns models of networked agents' behavior to predict their responses and reduce unnecessary communication and computation.
- Design learning-assisted adaptive quantization, communication and query methods that exploit correlations among agents, and validate them in applications such as sensor networks and smart grids.
More about the research
One protects privacy when sensitive medical videos are shared to train monitoring systems. The other cuts communication and computation when many networked agents optimize a shared goal.
Why it matters
Families of tracheostomy-dependent children could gain home monitoring systems trained on shared video without exposing identities; the same privacy tools may apply to daycare centers, nursing homes, hospices and prisons. Communication-efficient optimization could make sensor networks, cooperative robots, federated learning and smart power grids work with less bandwidth and computing power.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Guang Wang
Florida State University, FL·Computer Science
Build cross-domain data curation tools for MobilityNet that combine real mobility datasets from different sources to reduce data bias.
Sandeep Chinchali
University of Texas at Austin, TX·Engineering·Early-career
Develop collaborative inference algorithms that use a Markov Decision Process to decide whether and where a robot should offload computation to the cloud, given network delay, power and server load.
Miao Pan
University of Houston, TX·Computer Science
Design software-hardware co-designed model pruning and adaptive quantization schemes that exploit the sparsity of model updates to reduce computation and memory access during federated learning on a single 5G mobile device.
The school
Louisiana State University: 15 labs on PI Pages, $431M a year on research, 323 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.